Axel-Cyrille Ngonga Ngomo

MoE models slash the performance cost of multilingual expansion by 83% when scaling from 12 to 200 languages, fundamentally altering the trade-offs in capacity allocation.

Simplifying OWL class expressions can enhance reasoning efficiency by reducing complexity without sacrificing formal semantics.

Existing knowledge editing methods can accurately insert facts into LLMs but often fail to capture the broader logical implications, revealing a critical oversight in current evaluation practices.